Tampere University of Technology

TUTCRIS Research Portal

On requirements to accuracy of noise variance estimation in prediction of DCT-based filter efficiency

Research output: Contribution to journalArticleScientificpeer-review

Details

Original languageEnglish
Pages (from-to)139-154
Number of pages16
JournalTelecommunications and Radio Engineering
Volume75
Issue number2
DOIs
Publication statusPublished - 2016
Publication typeA1 Journal article-refereed

Abstract

Several approaches to prediction of denoising efficiency for the DCT-based filters with application to single and multichannel images corrupted by different types of noise have been proposed recently. For all of them, it has been assumed that the noise characteristics are a priori and accurately known. In this paper, we analyze the influence of errors in estimated noise variance on prediction accuracy. It is shown that this influence depends upon input parameter used in prediction. Based on the analysis, requirements to accuracy of the noise variance estimation are provided.

ASJC Scopus subject areas

Keywords

  • Additive noise, Denoising efficiency prediction, Estimation errors, Image filtering, Requirements to estimation accuracy

Publication forum classification